Combining Artificial Intelligence learning with WEARable devices for improved stress diagnostics

Wearable Biometric Monitoring Devices (WBMDs) represent a recent alterative for the collection of health-related data from people. WBMDs allow for continuous and real time data collection in natural settings. However, these devices face two key challenges: 1) WBMDs require seamless interfaces that are not traumatic to patients; 2) WDBMs continuous flow of data comes at the price of having an enormous amount of information that cannot be analyzed manually. While artificial intelligence (AI) systems emerge as a natural solution to deal with complex data, these require large sets of accurately annotated data, which are cannot be guaranteed. We will build new WBMDs in the form of skin-compliant electrodes, capable of recording electrophysiological signals while guaranteeing patients’ acceptance. Then, a novel AI system will be developed for improved neurodevelopmental stress assessment that exploits weakly annotated data from WBMDs jointly with traditional measures.